(co-authored with Jeremy Ferwerda)

Studies investigating whether refugee exposure motivates far-right voting have returned conflicting results. In this study, we argue that prior research has understated how electoral responses vary as a function of the spatial distance between voters and refugees. Focusing on Germany in the wake of the Syrian refugee crisis, we combine the street addresses of all refugee housing accommodations with ballot-box electoral returns. Using a difference-in-differences design within Berlin and Hamburg, we show that refugee exposure increased far-right voteshare by 3.6 percentage points among voters living in close proximity to refugee housing. However, these effects are hyper-local: on average, no effect can be detected among citizens living 300 meters or more from refugee accommodation. We further establish the generalizability of these findings across municipalities using census grid data. Our findings demonstrate how the geographic level at which treatment exposure is measured can affect the substantive conclusions of quasi-experimental designs.